U.S. patent number 11,022,568 [Application Number 16/740,685] was granted by the patent office on 2021-06-01 for method of determining the displacement of a component.
This patent grant is currently assigned to Rolls-Royce plc. The grantee listed for this patent is ROLLS-ROYCE plc. Invention is credited to Simon Cross, Akin Keskin, Luca Miller.
![](/patent/grant/11022568/US11022568-20210601-D00000.png)
![](/patent/grant/11022568/US11022568-20210601-D00001.png)
![](/patent/grant/11022568/US11022568-20210601-D00002.png)
![](/patent/grant/11022568/US11022568-20210601-D00003.png)
![](/patent/grant/11022568/US11022568-20210601-D00004.png)
![](/patent/grant/11022568/US11022568-20210601-D00005.png)
United States Patent |
11,022,568 |
Keskin , et al. |
June 1, 2021 |
Method of determining the displacement of a component
Abstract
A method of determining the displacement of a component within a
device during operation of the device, the method comprising the
steps of: obtaining a first x-ray image of the device while the
device is in a first operation state; obtaining a second x-ray
image of the device while the device is in a second operation state
different to the first operation state; processing each of the
first and the second image, wherein the processing comprises
applying a filter obtained based on the noise of the image and a
frequency characteristic of the image; superimposing the first and
the second images to align a predetermined point in each of the
first and the second images; and measuring the displacement of an
edge associated with the component between the first and the second
image to obtain the displacement of the component within the device
during operation of the device.
Inventors: |
Keskin; Akin (Derby,
GB), Miller; Luca (Derby, GB), Cross;
Simon (Derby, GB) |
Applicant: |
Name |
City |
State |
Country |
Type |
ROLLS-ROYCE plc |
London |
N/A |
GB |
|
|
Assignee: |
Rolls-Royce plc (London,
GB)
|
Family
ID: |
65997907 |
Appl.
No.: |
16/740,685 |
Filed: |
January 13, 2020 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20200240932 A1 |
Jul 30, 2020 |
|
Foreign Application Priority Data
|
|
|
|
|
Jan 30, 2019 [GB] |
|
|
1901244 |
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01N
23/04 (20130101); G06T 7/0004 (20130101); G01N
2291/2693 (20130101); G06T 2207/10124 (20130101); G01N
2291/2694 (20130101); G06T 2207/30164 (20130101); B64F
5/60 (20170101) |
Current International
Class: |
G01N
23/04 (20180101); B64F 5/60 (20170101) |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
1258924 |
|
Aug 1989 |
|
CA |
|
0234537 |
|
Sep 1987 |
|
EP |
|
0905509 |
|
Mar 1999 |
|
EP |
|
3255518 |
|
Dec 2017 |
|
EP |
|
Other References
European Search Opinion with Communication Transmittal for Patent
Application No. EP20150210.1 dated Mar. 5, 2020, 4 pages. cited by
applicant .
Great Britain search report dated Jul. 29, 2019, issued in GB
Patent Application No. 1901244.2. cited by applicant .
European search report dated Feb. 21, 2020, issued in EP Patent
application No. 20150210. cited by applicant .
Vardar, et al, "Failure Analysis of gas turbine blades in a thermal
power plant", Engineering Failure Analysis, vol. 14, No. , Jan. 18,
2007, pp. 743-749, 2006. cited by applicant .
KK Dhanya, "Image Registration Techniques Classification: A
Review", International Journal of Research in Advent Technology,
Jan. 1, 2014, pp. 2321-9637. cited by applicant.
|
Primary Examiner: Thomas; Courtney D
Attorney, Agent or Firm: Brinks Gilson & Lione
Claims
We claim:
1. A method of determining the displacement of a component within a
device during operation of the device, the method comprising the
steps of: obtaining a first x-ray image of the device while the
device is in a first operation state; obtaining a second x-ray
image of the device while the device is in a second operation state
different to the first operation state; processing each of the
first image and the second image, wherein the processing comprises
applying a filter obtained based on the noise of the image and a
frequency characteristic of the image or applying a transformation
to the image data to reduce the deviation of the frequencies of
appearance of pixel values within the image; superimposing the
first image and the second image to align a predetermined point in
each of the first image and the second image; and measuring the
displacement of an edge associated with the component between the
first image and the second image to obtain the displacement of the
component within the device during operation of the device.
2. The method of claim 1, wherein the processing of each of the
first image and the second image further comprises a normalisation
step comprising: determining the frequency of appearance of data
values within the image; and applying a transformation to the image
data such that the fraction of the transformed image data at least
one of a minimum and a maximum value has a predetermined value.
3. The method of claim 1, wherein the processing of each of the
first image and the second image further comprises: selecting a
pixel within the image; determining if the intensity value of the
pixel is above a predetermined threshold; and assigning an
intensity value to the pixel that is an average of the surrounding
pixels if the intensity value of the pixel is above the
predetermined threshold.
4. The method of claim 1, wherein the processing of each of the
first image and the second image further comprises a smoothing step
comprising: selecting a pixel within the image; and assigning an
intensity value to the pixel that is an average of the surrounding
pixels.
5. The method of claim 4, wherein the processing of each of the
first image and the second image comprises performing the smoothing
step before and after the filter is applied.
6. The method of claim 4, wherein the processing of each of the
first image and the second image comprises performing the
normalisation step before and after a smoothing step.
7. The method of claim 1, wherein the filter is derived to minimize
the mean square error between a frequency characteristic of an
ideal image and a frequency characteristic of the image.
8. The method of claim 1, wherein the filter is performed on plural
sub-regions of the image.
9. The method of claim 1, wherein the processing of each of the
first image and the second image further comprises an edge
detection step comprising at least one of the following: applying a
Gaussian filter to the image; calculating the image gradient and
each point in the image; suppressing non-maximum gradients pixels;
and performing hysteresis thresholding.
10. The method of claim 9, wherein the processing of each of the
first image and the second image further comprises removal of image
artefacts after the edge detection step.
11. The method of claim 1, wherein the first image and the second
image are aligned by detecting points of interest using a speeded
up robust features algorithm.
12. The method of claim 1, wherein the predetermined point is
associated with a predetermined component of the device.
13. The method of claim 1, wherein the device is a gas turbine
engine.
14. A method of determining the displacement of a component within
a device during operation of the device, the method comprising the
steps of: obtaining a first x-ray data set via a simulation, the
first x-ray data set corresponding to a first image of the device
while the device is in a first operation state; obtaining a second
x-ray image of the device while the device is in a second operation
state different to the first operation state; processing each of
the first image and the second image, wherein the processing
comprises applying a filter obtained based on the noise of the
image and a frequency characteristic of the image or applying a
transformation to the image data to reduce the deviation of the
frequencies of appearance of pixel values within the image;
superimposing the first and the second images to align a
predetermined point in each of the first image and the second
image; and measuring the displacement of an edge associated with
the component between the first image and the second image to
obtain the displacement of the component within the device during
operation of the device.
15. The method of claim 14, wherein the device is a gas turbine
engine.
16. A method of optimizing and/or monitoring a device, comprising
performing the method of claim 1, wherein the method further
comprises modifying the device or a component of the device or
planning a maintenance schedule for the device or the component of
the device based on the determination of the displacement of the
component.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
This specification is based upon and claims the benefit of priority
from United Kingdom patent application number GB 1901244.2 filed on
Jan. 30, 2019, the entire contents of which are incorporated herein
by reference.
BACKGROUND
Technical Field
The present disclosure relates to a method of determining the
displacement of a component within a device during operation of the
device, a computer program comprising code configured to instruct
the computer system to perform the method and a computer system for
determining the displacement of a component within a device during
operation of the device.
Description of the Related Art
X-ray images are a useful means of understanding detailed device
behaviour during the design, operation and verification of devices
such as gas turbine engines. X-ray images may be obtained at
different locations within a device and under different operating
conditions of the device. Such images may then be analysed to
understand if the device is behaving as intended under the
different operating conditions of the device. In particular, the
location of particular components in different operating states of
the device may be obtained. Capture, analysis and storage of the
images may be performed digitally.
One approach is to use the x-ray image data and the assistance of a
human operator to manually superimpose the x-ray images of the
device in different modes of operation. The superposition of the
images allows the operator to visualise the displacement of
components within the device during the different operation states.
The movement of the component may further be derived based on
displacement relative to the different modes of operation.
However, the x-ray imaging and analysis process can take a
significant amount of time. In addition to this, due to the nature
of manual superposition, different operators may derive different
estimated locations of a component. In addition to this, the
accuracy of the estimated location of the component can vary
between operators. The accuracy of the determination may be
increased by averaging the estimates obtained from multiple human
operators. However, this requires multiple operators to review the
image and extends the time required to perform the analysis even
further.
It can therefore be seen that there is a desire for a method of
determining the displacement of a component within a device that is
capable of providing accurate measurements in a short amount of
time.
SUMMARY
According to a first aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray image of the device while the device is in
a first operation state; obtaining a second x-ray image of the
device while the device is in a second operation state different to
the first operation state; processing each of the first image and
the second image, wherein the processing comprises applying a
filter obtained based on the noise of the image and a frequency
characteristic of the image or applying a transformation to the
image data to reduce the deviation of the frequencies of appearance
of pixel values within the image; superimposing the first image and
the second image to align a predetermined point in each of the
first image and the second image; and measuring the displacement of
an edge associated with the component between the first image and
the second image to obtain the displacement of the component within
the device during operation of the device.
According to a second aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray image of the device while the device is in
a first operation state; obtaining a second x-ray image of the
device while the device is in a second operation state different to
the first operation state; processing each of the first image and
the second image, wherein the processing comprises applying a
filter obtained based on the noise of the image and a frequency
characteristic of the image; superimposing the first image and the
second image to align a predetermined point in each of the first
image and the second image; and measuring the displacement of an
edge associated with the component between the first image and the
second image to obtain the displacement of the component within the
device during operation of the device.
According to a third aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray image of the device while the device is in
a first operation state; obtaining a second x-ray image of the
device while the device is in a second operation state different to
the first operation state; processing each of the first image and
the second image, wherein the processing comprises applying a
transformation to the image data to reduce the deviation of the
frequencies of appearance of pixel values within the image;
superimposing the first image and the second image to align a
predetermined point in each of the first image and the second
image; and measuring the displacement of an edge associated with
the component between the first image and the second image to
obtain the displacement of the component within the device during
operation of the device.
Optionally the processing of each of the first image and the second
image further comprises a normalisation step comprising:
determining the frequency of appearance of data values within the
image; and applying a transformation to the image data such that
the fraction of the transformed image data at least one of a
minimum and a maximum value has a predetermined value.
Optionally, the processing of each of the first image and the
second image further comprises: selecting a pixel within the image;
determining if the intensity value of the pixel is above a
predetermined threshold; and assigning an intensity value to the
pixel that is an average of the surrounding pixels if the intensity
value of the pixel is above the predetermined threshold.
Optionally, the processing of each of the first image and the
second image further comprises a smoothing step comprising:
selecting a pixel within the image; and assigning an intensity
value to the pixel that is an average of the surrounding
pixels.
Optionally, the processing of each of the first image and the
second image comprises performing the smoothing step before and
after the filter is applied. Optionally, the processing of each of
the first image and the second image comprises performing the
normalisation step before and after a smoothing step.
Optionally, the filter is derived to minimize the mean square error
between a frequency characteristic of an ideal image and a
frequency characteristic of the image.
Optionally, the filter is performed on plural sub-regions of the
image.
Optionally, wherein the processing of each of the first image and
the second image further comprises an edge detection step
comprising at least one of: applying a Gaussian filter to the
image; calculating the image gradient and each point in the image;
suppressing non-maximum gradients pixels; and performing hysteresis
thresholding.
Optionally, the processing of each of the first image and the
second image further comprises removal of image artefacts after the
edge detection step. Optionally the images are aligned by detecting
points of interest using a speeded up robust features
algorithm.
Optionally, the predetermined point is associated with a
predetermined component of the device.
According to a fourth aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray data set via a simulation, the first x-ray
data set corresponding to a first image of the device while the
device is in a first operation state; obtaining a second x-ray
image of the device while the device is in a second operation state
different to the first operation state; processing each of the
first image and the second image, wherein the processing comprises
applying a filter obtained based on the noise of the image and a
frequency characteristic of the image or applying a transformation
to the image data to reduce the deviation of the frequencies of
appearance of pixel values within the image; superimposing the
first image and the second image to align a predetermined point in
each of the first image and the second image; and measuring the
displacement of an edge associated with the component between the
first image and the second image to obtain the displacement of the
component within the device during operation of the device.
According to a fifth aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray data set via a simulation, the first x-ray
data set corresponding to a first image of the device while the
device is in a first operation state; obtaining a second x-ray
image of the device while the device is in a second operation state
different to the first operation state; processing each of the
first image and the second image, wherein the processing comprises
applying a filter obtained based on the noise of the image and a
frequency characteristic of the image; superimposing the first
image and the second image to align a predetermined point in each
of the first image and the second image; and measuring the
displacement of an edge associated with the component between the
first image and the second image to obtain the displacement of the
component within the device during operation of the device.
According to a sixth aspect there is provided a method of
determining the displacement of a component within a device during
operation of the device, the method comprising the steps of:
obtaining a first x-ray data set via a simulation, the first x-ray
data set corresponding to a first image of the device while the
device is in a first operation state; obtaining a second x-ray
image of the device while the device is in a second operation state
different to the first operation state; processing each of the
first image and the second image, wherein the processing comprises
applying a transformation to the image data to reduce the deviation
of the frequencies of appearance of pixel values within the image;
superimposing the first image and the second image to align a
predetermined point in each of the first image and the second
image; and measuring the displacement of an edge associated with
the component between the first image and the second image to
obtain the displacement of the component within the device during
operation of the device.
Optionally, the fourth, fifth or sixth aspect is combined with at
least one of the optional features discussed above.
Optionally, the device is a gas turbine engine.
According to a seventh aspect there is provided a computer program
comprising code means that, when executed by a computer system,
instructs the computer system to perform the first to sixth
aspect.
According to a eighth aspect there is provided a computer system
for determining the displacement of a component within a device
during operation of the device, the system comprising at least one
processor and memory, the memory storing code that performs the
method of the first to sixth aspect.
According to a ninth aspect there is provided a method of
optimizing and/or monitoring a device, comprising performing the
method of the first to sixth aspect, wherein the method further
comprises modifying the device or a component of the device or
planning a maintenance schedule for the device or the component of
the device based on the determination of the displacement of the
component.
DESCRIPTION OF THE DRAWINGS
Embodiments will now be described by way of example only, with
reference to the figures in which:
FIG. 1 is an example of images at different processing steps.
FIG. 2 is a flow chart illustrating an example set of process steps
for processing an image.
FIG. 3 is a flow chart illustrating an example set of process steps
for comparison of images.
FIG. 4 is a flow chart illustrating an example set of process steps
for conducting measurements on an image.
FIG. 5 is a flow chart illustrating an alternative example set of
process steps for processing an image.
DETAILED DESCRIPTION
Aspects and embodiments of the present disclosure will now be
discussed with reference to the accompanying figures. Further
aspects and embodiments will be apparent to those skilled in the
art.
FIG. 1 illustrates examples of images at different stages of
processing. First and second images 1,2 are separate images of a
device that have been obtained of the device in different operating
conditions.
Each of the first and second image 1,2 may be obtained using x-ray
imagery of the device. The images may therefore indicate the
positions of components within the device that are not visible by
physical inspection of the device. The intensity of the pixels
within the images may indicate the position of material within the
device that forms the various components of a device. For example,
dark pixels associated with a lower intensity in the image may
indicate the presence of material, whereas light pixels associated
with a higher intensity may indicate the lack of presence of a
material. Areas of dark pixels associated with a lower intensity
may therefore indicate the presence of a component in the image.
The boundary between an area of dark pixels associated with a lower
intensity and an area of light pixels associated with a higher
intensity may therefore be the edge of a component in the
device.
In the example discussed above, the images 1,2 are obtained using
x-ray imagery of the device. However, the images may supplemented
by image data obtained by another imaging method such as ultrasound
imagery, visible light imagery, thermal imaging, acoustic imagery
or any other technique capable of producing special information
related to the materials forming the device.
At least one of the images 1,2 may be obtained by a simulation of
the device. For example, a mechanical simulation involving
simulation of each component of the device and the interactions of
these components may be used to produce a 3D computer model of the
device in various states of operation. Images of sections of the
device obtained from the 3D computer model may be compared to each
other using the method described in detail below. Images of
sections of the device obtained from the 3D computer model may also
be compared to images obtained by imaging of a physical device.
The images 1,2 may be obtained in varying states of operation of
the device. For example, at least one of the images 1,2 may be
obtained when the device is not operating. Such a state of
operation may be referred to as a cold state because the components
of the device have not increased in temperature due to the
operation of the device. Further operation states of the device may
be when the device has been in operation for a certain amount of
time. A further state of operation of the device may include a
state in which the device is not operating, but immediately after
the device has been operating for a period of time. Further states
of operation may be when different components within the device are
in operation. For example, different air systems in the device may
be operating in different states of operation of the device.
The components of the device may also move within the device for
other reasons. For example, operation of the device may result in
the temperature of the components of the device increasing and the
position of the components therefore shifting within the device.
Components may shift due to vibration of the device during
operation of the device. Components may also move within the device
due to the intended operation of the device. For example, values or
switches may open and close during operation of the device.
The images 1,2 may then undergo image processing to produce
processed images 3,4. In the example shown in FIG. 1, the first
image 1 has been processed to obtain a first processed image 3. The
second image 2 has been processed to obtain a second processed
image 4. The image processing process is described in greater
detail below. Following the image processing process, the first and
second processed images 3,4 may then be combined to form a combined
image 5. The image combining process is described in greater detail
below. The combined image 5 may then be analysed to determine the
displacement of the component in the device. This analysis is
described in greater detail below.
FIG. 2 shows an example of the image processing process that may be
applied to obtain the processed images 3,4 using the initial images
(1,2). In a first step 10, an input image 1,2 may be received. The
input image may have been obtained using any of the methods
discussed above. The input image may, for example, be a grey scale
or colour representation of the x-ray data obtained by x-ray
imaging of the device. The input image may, for example, be stored
in a 1-bit, 8-bit or 16-bit greyscale format.
In a second step 11, the data in the image may be normalised. The
normalisation of the data may be based on the frequency of
appearance of the data. The normalisation of the data may involve
identifying pixel values or pixels that fall within at least one of
the bottom 5% and or top 5% of the frequency of appearance,
optionally at least one of the bottom 1% or top 1% of the frequency
of appearance, and replacing these pixels with a replacement data
value. The normalisation of the data may involve identifying pixel
values or pixels that fall within at least one of the bottom 5% and
or top 5% of the intensity values of the image, optionally at least
one of the bottom 1% or top 1% of the intensity values of the
image, and replacing these pixels with the replacement data value.
The replacement data value may be an average of the pixels
surrounding the selected pixel. The replacement data value may be a
minimum value (also known as minimum saturation) of intensity for
an identified pixel falling with the bottom 5% or 1% of intensity
values. The replacement data value may be a maximum value of
intensity (also known as maximum saturation) for a identified pixel
falling with the top 5% or 1% of intensity values. The
normalisation of the data of the image may involve applying a
transformation to the image data such that a predetermined quantity
of the intensity values of the pixels of the transformed image data
are at least one of a minimum or a maximum saturation. The
predetermined quantity may be, for example, 5% or 1% of the total
number of pixels in the image. The data values of the pixels may
further be normalised to a value between 0 and 1 during this
step.
The normalisation step may introduce localised noise into the image
data. The localised noise may be referred to as noise spikes in the
data. In a third step 12, spikes of noise may be located and
removed from the data. For example, the image data may be analysed
to determine pixels that have an intensity value that is over a
predetermined threshold. Pixels that have an intensity value over
the predetermined threshold may be determined to contain noise. The
data value in pixels determined to be containing noise may be
replaced with a value that is an average of the surrounding pixels.
Alternatively, the data value in the pixels may be replaced with a
predetermined pixel value. Both the predetermined threshold and the
predetermined pixel value, where used, may be set based on previous
analysis of the images.
In a fourth step 13, the data in the image may be smoothed.
Smoothing may be performed by applying a nearest neighbours
algorithm to the image. In the application of a nearest neighbours
algorithm, a pixel is selected and the data value of that pixel is
replaced by an average of the data value of the pixel and a number
of surrounding pixels. The surroundings pixels may include each of
the eight pixels immediately adjacent to the selected pixel. In
this case, the nearest neighbours algorithm is being applied using
a 3.times.3 mask. The nearest neighbours algorithm may use a larger
mask, in which case an average data value of all of the pixels
located under the mask is used to replace the data value of the
selected pixel. The weighting of the pixels may vary. For example,
the pixel distribution may be weighted using a Gaussian function
when obtaining the average value of the pixels under the mask.
In a fifth step 14, a filter may be applied to the image. The
filter may be 2D adaptive noise removal filter. The filter may be
obtained based on the noise of the image and a frequency
characteristic of the image. The frequency characteristic of the
image may be obtained by performing a Fourier transform on the
image data. The filter may be a Wiener filter and may therefore
minimize the mean square error between the frequency characteristic
of the image before the noise was applied to the image and the
obtained image. The filter may, for example, be applied using a
9.times.9 mask. Other mask sizes may be used.
In a sixth step 15, a further smoothing step may be applied. The
further smoothing step may be performed using the same conditions
as the fourth step 13. However, the conditions may vary. For
example, a different size mask or a different weighting of the
pixel values within the mask may be used.
In a seventh step 16, an edge detection process may be applied. The
edge detection process may result in the removal of data from the
image that are determined not to form part of an edge in the image
from the image. Pixels corresponding to data that is not determined
to form part of an edge may be set to a zero value. The edge
detection process may be a canny edge detector.
The edge detection process may comprise at least one of the
following sub-steps. A first sub-step of the edge detection process
may be an application of a Gaussian filter to the image. For
example, the image data may be convoluted with a Gaussian matrix. A
second sub-step of the edge detection process may be the
determination of the intensity gradient of each pixel of the image.
A third sub-step of the edge detection process may be the
application of non-maximum suppression. In this step, the gradient
of each pixel may be compared to adjacent pixels and the pixel may
be determined to be part of an edge if the gradient value is larger
than the adjacent pixels.
A fourth sub-step of the edge detection process may be a hysteresis
step. In the hysteresis step, an upper and a lower threshold may be
selected. Pixel gradients which are higher than the upper threshold
may be determined to be part of an edge. Pixel gradients which are
lower than the lower threshold may be determined not to be part of
an edge. The lower threshold, for example may be
7.5.times.10.sup.-2 in the case where the image data has been
normalised to a maximum value of 1. Pixel gradients which are
between the upper and the lower threshold may be determined to be
part of an edge if the pixel is adjacent to a pixel above the upper
threshold and may be determined not to be part of an edge if this
is not the case.
In an eighth step 17, image artefacts may be removed from the
image. The image artefacts removed may have been introduced due to
the edge detection process performed in the seventh step 16. The
image artefacts may be located by determining an area of pixels
with a particular high or low intensity value with a total area
that is less than a predetermined value. The predetermined value
may be, for example, 100 pixels. The pixels determined to be part
of the artefact may be replaced with an average pixel value of the
image or a zero value.
In a ninth step 18, the processed images 3,4 may be output for
analysis.
The processing steps may be performed in the order set out above.
However, the method is not restricted to this order and some of the
steps may be omitted and performed at a different time. Any of the
steps may be repeated more than once. For example, only the
filtering step may be performed. In a further example, only one of
the smoothing step 13 and further smoothing step 15 may be
performed. The steps above are described in relation to individual
pixels of the image. Any of the steps described above may be
performed on each of the pixels in the image. Any of the steps
described above may be performed on at least one sub region of the
image.
The combining process will now be described in more detail. An
example of the steps that may be performed in the combining process
is shown in FIG. 3. In a first combining step 20, two processed
images 3,4 are selected for comparison. The images may both have
been processed using the image processing process described
above.
In a second combining step 21, features within the processed images
3,4 are matched to align the two images. Even if each of the
processed images 3,4 were originally obtained from an identical
perspective, displacement of the device within the images may have
occurred between the two images being obtained. For example, in the
case where the device is a gas turbine engine, displacement may
have occurred due to engine mount movement during operation of the
device. The processed images 3,4 may therefore be aligned based on
a point associated with a particular component within the device.
This component may be a predetermined component which is known to
move minimally between the different operating conditions of the
two images.
The processed images 3,4 may initially be aligned using a
speeded-up robust features (SURF) algorithm to automatically detect
features within the edge data. The SURF algorithm is used to
pre-align the images to decrease later work required. The algorithm
may involve the sub-steps of determining a notable feature or
keypoint in the first image, determining the orientation of the
feature or keypoint and determining a descriptor containing
information of the neighbourhood of the image of the keypoint. The
determination of the keypoint may be performed using blob
detection. The descriptor may then be used to orientate the first
image with the second image. The processed images 3,4 may be
initially aligned by a manual operator. The alignment may be
performed based on a sub-region of at least one of the images. The
alignment process may generate a 2D displacement vector
corresponding to the transformation required to align the processed
images 3,4.
In a third combining step 22, the alignment of the images may be
improved by the manual operator using a graphical user interface
(GUI). The GUI may display the images superimposed on one another
to the manual operator. The user may be automatically directed to
click on two points in the image that correspond. The displacement
between the two points may then be used to improve the alignment of
the images. The alignment step may be repeated multiple times until
the images are aligned acceptably, resulting in a combined image
5.
The analysis process will now be described in more detail. An
example of the analysis process is shown in FIG. 4. In a first
analysis step 30, a combined image 5 is selected for measurement
and loaded into a GUI.
In a second analysis step 31, the user may click two points within
the image that correspond to the same component, and/or the same
part of a component in the different operating steps of the device.
In a third analysis step 32, the displacement between the two
points may then be calculated. The displacement may be calculated
using a scaling factor. The scaling factor may be predetermined by
the interface. Alternatively, the scaling factor may be determined
by the user. Alternatively, the scaling factor may be embedded in
the combined image 5. The displacement value may be displayed to
the user. The displacement value may be scaled based on the device
that is present in the combined image 5. For example, in the case
of a gas turbine engine, the axial and radial displacement between
the two points may be calculated.
The analysis process may be performed multiple times by different
users and the result averaged to produce a final displacement
measurement.
The method as described above may be applied to any device which
may contain components which may become displaced in different
operating states. For example, the method may be applied to a gas
turbine engine. The method may be applied to other parts of an
aeroplane, such as wings, fuselage or other devices, or components
within an engine such as manifolds or channels.
The device being imaged may further be modified based on the
measured displacement results. For example, if a component of the
device is identified to be moving in an undesirable way between the
different operating states of the device, the component may be
modified. For example, if the movement is due to the high
temperature of the device during an operating step, the material of
the component may be modified based on the measured displacement to
reduce the displacement. The location of components within the
device may also be varied depending on the displacement measured.
In addition to this, maintenance data based on the displacement of
components may be determined based on the size of the displacement
of the component during the operation of the device and a
maintenance regime including, for example, an operation lifetime of
a component, may be calculated.
In an alternative embodiment, the image processing process may
comprise a transformation step. An example of an image processing
process including a transformation step is shown in FIG. 5. In a
first step 50, an input mage process may be received. In a second
step 51, a smoothing process may be performed. In a third step 52,
a normalisation process may be performed. Each of these processes
may correspond to the processes performed in the image processing
process described above.
In a fourth step 53, an image transformation process may be
performed. The transformation may reduce the deviation of the
frequencies of appearance of pixel values within the image. The
frequencies of appearance of pixel values may be determined by
dividing the range of possible intensity values of the pixel data
into a plurality of sub-ranges and assigning each pixel to one of
the plurality of sub-ranges to determine the frequencies of
appearance of the pixel values. The transformation may modify the
intensity value of at least one of the pixels such that the
deviation of the frequencies of appearance of the pixel values is
reduced. The deviation may be a statistical measure of the variance
of the frequencies of appearance of the pixel values. For example,
the deviation may be the standard deviation of the frequencies of
appearance of the pixel values. The transformation may equalize the
frequency of appearance of each of the plurality of sub-ranges of
pixel values. The transformation may be a histogram equalisation of
the image data.
In a fifth step 54, a further normalisation process may be
performed. In a sixth step 55, the processed image may be output
for further analysis as described above. Each of these processes
may correspond to the processes performed in the image processing
process described above.
The processing steps in the alternative embodiment may be performed
in the order set out above. However, the method is not restricted
to this order and some of the steps may be omitted and performed at
a different time. Any of the steps may be performed more than once.
For example, only the transformation process may be performed. In a
further alternative, the transformation process may be performed
instead of the step of applying the filter described in the first
embodiment. Alternatively, the step of applying the filter
described in the first embodiment may be performed instead of the
transformation process in the alternative embodiment discussed
above. The steps above are described in relation to individual
pixels of the image. Any of the steps described above may be
performed on each of the pixels in the image. Any of the steps
described above may be performed on at least one sub region of the
image.
It will be understood that the invention is not limited to the
embodiments above described. Various modifications and improvements
can be made without departing from the concepts described herein.
Except when mutually exclusive, any of the features may be employed
separately or in combination with any other features and the
disclosure extends to and includes all combinations and
sub-combinations of one or more features described herein.
* * * * *